EXPERIMENTAL STUDY OF PIANO KEY WEIR

Abstract

As projects are reassessed for safety, provision for an increased estimate of the probable maximum flood (PMF) has to be made in many cases. It is therefore necessary to provide more flood storage and/or larger capacity for spillways to pass the PMF safely. An innovative and effective way of increasing the spillway capacity is to use a labyrinth weir. The concept of the Labyrinth weir is to vary the plan shape of the crest to increase the effective crest length. This increases the discharge per unit width of the spillway for a given operating head. A Labyrinth weir has advantages compared to the straight over flow weir and the standard ogee crest Labyrinth weirs can be used to increase outlet capacity for a given spillway crest elevation and length or to increase storage by raising the crest while maintaining spillway capacity. A new concept of a labyrinth weir has been proposed with a new shape like black and white Piano keys when viewed in plan. Keeping in view the relatively reported better performance of Piano Key Weirs in comparison to linear and Labyrinth weirs, following objectives are considered for the present study: (i) to perform experiments on different configurations of Piano Key Weir, varying in term of presence/absence of ramp in upstream and downstream for one side and both side overhanging and varying ratio of any two consecutive limbs, (ii) to identify the situations which lead to increase in the magnification ratio of discharge passing through the Piano Key Weir, (iii) to quantify the maximum achievable magnification ratio and associated hydraulic conditions, (iv) to identify the parameters influencing the variation of discharge coefficient and study the role of considering the total head versus the head over the crest and (v) to develop a set of relationships for the variation of discharge coefficient and to identify the most suitable relationship The work includes details of five phases of experiments. Initial phase of experiments (phase -I to IV) which were planned with different configurations of Piano Key Weir, i.e. with or without ramp and one or two side overhanging (u/s & d/s) indicate that Piano Key Weir with presence of ramp and two sides overhanging (u/s & d/s) provides a higher discharge under same head when compared with other Piano Key Weir configurations with lesser number of ramps and/or over-hangings. In phase-V experiments, filling were introduced in the ramps but these were not found to increase the discharge. Thus, ramps with no planar discontinuity were found to be best performing. In phase-V experiments, also modification were introduced into inlet limb of Piano Key i i Weir but it was again observed that such an inlet modification was of no practical significance as it did not lead to any increase in the discharge. The ratio (r) of Piano Key Weir to linear weir discharge for a given head was always more than one and when compared with Labyrinth weir (based on computational only) was always higher than the Labyrinth weir. This finding is in conformity with the literature. The ratio (r) was found to increase with magnification ratio L/W. However, at larger value of L/W, the ratio (r) was observed to tend to approach a limiting value in the proximity of four. For a very large value of L/W, where L is the perimeter of the Piano-Key Weir crest and W is the width of the channel, it was observed that variation of ratio of inlet and outlet cell width did influence the ratio r and the performance was best when two cells were of same width. Any deviation from this ratio was found to have a negative effect on ratio r at larger L/W ratios. Among the several options attempted to develop relationship for Cd variations for two different configurations of Piano Key Weir, i.e. one side and both side overhanging, it was found that for one side overhanging Piano Key Weir, Cd variations as a function of (i) Fronde number, Fr (ii) h/p (head over the weir, h and weir height asp) and (iii) Fr and h/p leads to development of several relationships with a relatively higher values of statistical fits, i.e. Coefficient of Determination (R2) indicating the appropriateness of different functional relationships. However, the perusal of function relationships for both side overhanging indicates relatively poor performance of Cd variation only as function of Fr. Also, the Cd variation with h/p as only variational parameter doesn't appear to work so well (R2 = 0.80) when compared with counterpart for one side overhanging (R2 = 0.96). The use of total head is also explored in the analysis of Cd variations. However, no significant improvements are observed into development of Cd relationships. This may be because of the reason that velocity heads are very small in the experiments performed. Study also indicates that for lower value of L/W, say as 3.56 and 4.84, the functional relationship of Cd variation in term of Fr and h/p appears to be most suitable choice. However, for larger value of L/W as 7.4, relationship is strongly dominated by h/p. Evaluation of the data collected at field scale model using (i) Cd = f(Fr), (ii) Cd = f(114) (iii) Cd =ifir/p, Fr) indicates that the use of Cd = f(hip, Fr) very well works with the field scale model results. Ill